This project will restore arm and hand function to individuals with complete paralysis using functional electrical stimulation (FES) AND will give these people the ability to command these movements in an effective and intuitive way using an intracortical brain-machine interface (BMI). We will use percutaneous interfaces for both the FES and BMI components to implement a fully functional but also reversible BMI-commanded FES system. This is the immediate next step in the ultimate realization of a permanently implanted BMI-controlled FES system. Paralyzed muscles of 5 individuals with high level (C1-C4) spinal cord injuries will be implanted with FES electrodes to restore multiple motions of the arm and hand sufficient for meaningful multi-joint, functional activities. In the same individuals, a 96-channel intracortical array (BrainGate2) will be implanted in the arm/hand area of primary motor cortex, and the resulting signals will be used to command the motions of the participant's arm and hand via thought. The main components of the proposed system are: intramuscular electrodes with percutaneous leads (Ardiem Medical), an external stimulator (FES Center), a BrainGate2 intracortical array and associated external hardware (Neuroport by Blackrock Microsystems), and a standard computer with a real-time operating system (Matlab xPC Target) as the FES controller. Participants will be strongly motivated to optimize the performance of a fully functional system that drives their own paralyzed arms, and they will be given ample opportunity to practice and learn the interfaces. We will test the control performance for three different command interfaces that have been widely used in similar applications: (1) continuous trajectory control used widely in previous BMI research, (2) movement goal-based control widely used to control robotic arms, and (3) state-based gated ramp control used widely to control artificial prosthetic arms. Participants will perform the same set of standard movements as well as functional activities with each interface. We will compare the effectiveness and robustness of each command approach based on technical and functional performance metrics (accuracy, speed, consistency over time, functional performance, ease of use). We will also evaluate the ability of M1 to generate continuous, goal and state commands, and will characterize changes in neural signal properties (tuning and modulation depth) while using these three interfaces. This project will, for the first time, directly test the feasibility of a human intracortical BMI-controlled FES upper limb system, so our results will guide the specifications of future, fully-implanted BMI systems. Our team has 30+ years of experience in developing and testing upper limb FES systems, including in people with complete arm paralysis. We have been working to develop a human intracortical BMI for the past 7 years, have full regulatory approval, and have established a clinical BrainGate2 site in Cleveland. This project is a natural expansion of our past work by combining the FES and BMI approaches in people with SCI. The technical risks of this project are relatively low, but the potential scientific and rehabilitation returns are very high.